"ATi & nVidia Anti-Aliasing Performance"

So what do they conclude? It looks like they tried to make comparisons based on equivalent qualities -- which certainly is good. 4x on ATI, for example, is alot better looking than 4x on NVidia. My own opinion based on screenshots has been that a NVidia card must use at least 6xS to match ATI's 4x.
 
Leonidas set forth to compare Antialiasing quality. Unfortunately he did so with AF turned on, oh well ...

The German audience is in an outrage, as usual :D
A translation is in the works.
 
Problem is, as i've posted in the 3DC-Forums, that the author of this article is trying to compare equal anti-aliasing quality, as zeckensack already mentioned.

To make it a more realistic comparison he added AF to the settings but judged the resulting quality solely by the EER, disregarding that due to the supersampling part in nVs horribly slow hybrid modes, the textures get a decent boost too.

That makes this article especially with regard to its title a little hard to interpret correclty IMO.
 
I didn't bother translating...but looking at the article quickly.... How the hell is anyone supposed to compare quality between the two cards when they are on seperate pages? I would have much preferred to see the modes side by side to compare the different modes.
 
Razor04 said:
I didn't bother translating...but looking at the article quickly.... How the hell is anyone supposed to compare quality between the two cards when they are on seperate pages? I would have much preferred to see the modes side by side to compare the different modes.
You can pick the modes you want and open up the large images in background pages (the inline images are only thumbnails). If that doesn't work for you, try using a decent browser ;)
 
zeckensack said:
Razor04 said:
I didn't bother translating...but looking at the article quickly.... How the hell is anyone supposed to compare quality between the two cards when they are on seperate pages? I would have much preferred to see the modes side by side to compare the different modes.
You can pick the modes you want and open up the large images in background pages (the inline images are only thumbnails). If that doesn't work for you, try using a decent browser ;)
hehe woops...didn't see that but then again I didn't pay much attention cause I don't read German. :) Oh and last I checked Firebird was a pretty good browser. ;)
 
Ostsol said:
So what do they conclude? It looks like they tried to make comparisons based on equivalent qualities -- which certainly is good. 4x on ATI, for example, is alot better looking than 4x on NVidia. My own opinion based on screenshots has been that a NVidia card must use at least 6xS to match ATI's 4x.

Concentrated on edge quality alone, 6xS according to my realtime experience is only slightly better than 4xS (3*4 vs 2*4).

Also for those that don't understand german, 6xS has the tendency to create artifacts at times:

Ebenfalls eher zweifelhaft ist der 6xS Modus, da dieser zur Artefakt-Bildung tendiert. Den Modus 6x hat nVidia im übrigen genauso aus den 50er Treibern entfernt, geblieben ist nur der - Artefakt-behaftete Modus - 6xS, im nVidia Control Panel allerdings "6x" genannt.
 
Is it correct that this article is based only on edge AA quality? There are a lot of situations where 4xS looks much better than ATI 4X due to the supersampling approach, especialy in motion. Comanche4 is a good example for it.
 
Unless the scenery is overloaded with alphas, 4xS is not by far superior than a 4x sparse/ high degree AF combination. Even then the supersampling in that specific case gets applied only on one axis. ATI's AF algorithm does filter textures on all angles with a minimum of 2x samples at least.

4xS + 2xAF is of course another story, yet the advantage remains mostly for alphas.
 
Ailuros said:
Unless the scenery is overloaded with alphas, 4xS is not by far superior than a 4x sparse/ high degree AF combination.

IMO 4xS is better in a lot of games, even if they do not use a lot of alpha. The picture feels much "smoother" with SS enabled. High degree AF without SS often leads to strange texture "wobbling". It's a pitty I can't prove that unless you directly see it in motion side by side. :?

Even then the supersampling in that specific case gets applied only on one axis. ATI's AF algorithm does filter textures on all angles with a minimum of 2x samples at least.

At least 4xS does supersample on any surface. AF algorithms do only "supersample" surfaces with streched textures.
 
Mephisto said:
Ailuros said:
Unless the scenery is overloaded with alphas, 4xS is not by far superior than a 4x sparse/ high degree AF combination.

IMO 4xS is better in a lot of games, even if they do not use a lot of alpha. The picture feels much "smoother" with SS enabled. High degree AF without SS often leads to strange texture "wobbling". It's a pitty I can't prove that unless you directly see it in motion side by side. :?
I'll agree with that. . . Supersampling does wonders for texture quality. :)
 
IMO 4xS is better in a lot of games, even if they do not use a lot of alpha. The picture feels much "smoother" with SS enabled. High degree AF without SS often leads to strange texture "wobbling". It's a pitty I can't prove that unless you directly see it in motion side by side.

It's a pity that I've played with both alternative sollutions for more than just a couple of days I'd say. In the majority of cases a higher resolution with a MSAA/AF combination cures most problems apart from alphas. Any form of Supersampling will restrict to specific resolutions only due to the fillrate impact.

Try F1 2002 with 4xS in let's say 1024*768*32 vs. 4x sparse/16xAF in 1280*960*32 or higher. Supersampling is NOT the ultimate cure for everything. A SSAA + high degree of Anisotropic though is, but I'd love to see the hardware that is capable of high quality SSAA/AF combinations in high resolutions.

At least 4xS does supersample on any surface. AF algorithms do only "supersample" surfaces with streched textures.

It's still just ONE axis that you'll get supersampling on with 4xS. Would we be talking about THE ultimate sollution present in todays hardware it would be 8xS (2x sparse + 4xOGSS) combined with 8xAF. But since resolutions and performance matter to me at least too, it's not that usable after all apart from some rare extremely CPU bound corner cases. High resolutions are out of the question too.
 
Back
Top